Chillin'Competition

Nicolas Petit’s book Big Tech and the Digital Economy comes at a time of increasingly polarised and politicised debates on how one solves a “problem” like big tech.[1] It presents a meticulously studied anatomy of six big tech companies – Google, Facebook, Amazon, Apple, Microsoft, and Netflix. It offers nuanced prescriptions of what one can do about this group of companies, which displays levels of growth, size, and market capitalisation that are unheard of in economic history. Petit’s proposals are primarily aimed at competition law and regulation, but the book has a wide scope that will be of appeal to a broader audience due to the book’s incorporation of management, technology, economics, and different areas of law such as tax and data protection.

Petit’s book easily distinguishes itself from the crowd by its clear-headed inquiry into whether there is a problem with big tech and, if so, what competition law can and should do about it. His analysis is objective, yet critical, unlike the – what Petit calls – “airport books” on the topic, where everyone can find a narrative about what’s right or wrong with big tech to suit her personal taste or “echo chamber”. Petit’s book, in contrast, displays the best qualities of public discourse on a subject of contemporary significance (such as open-mindedness), whilst avoiding the worst (such as the ad hominem). It is both very accessible for the non-specialist and full of thought-provoking ideas for the specialist. This particular piece will discuss one such important and interesting theory put forward in the book, namely that traditional antitrust principles modelled on “rivalry” should be abandoned or radically altered in digital markets.[2]

According to Petit, the protection of rivalry is not always socially beneficial in industries with increasing returns to adoption, such as the digital markets under scrutiny.[3] Petit’s theory is premised on the idea that “in industries with increasing returns, economic forces … produce incentives on structural monopoly firms to compete by indirect entry in untipped markets, and avoid privately and socially inefficient rivalry in tipped markets”.[4] According to Petit, “[r]ivalry in tipped markets is privately inefficient because it is very costly for the direct entrant, and it is socially inefficient when there are increasing returns to scale due to rising marginal benefits”.[5] Thus, “[a] sound antitrust regime should … preserve competitive pressure on monopoly firms in markets that have tipped and in which incentives to indirect entry have therefore disappeared”.[6] One implication is that “antitrust should focus on cases of harm to competition in markets that have tipped, and be more forgiving toward the leveraging of market power in untipped markets”.[7]

Petit’s proposition that antitrust should abandon its principles based on rivalry when it is applied in digital markets is controversial, because, taken out of context, it may sound like giving up on competition. Yet, it has intuitive appeal, within the context in which Petit frames it, namely on the basis of the distinction that Petit convincingly makes between tipped and untipped digital markets. In essence, the theory is built upon the observation of how competition between the tech giants actually occurs in digital markets. It tries to incorporate those facts of competition within a theory of “moligopoly”. The theory explains that the tech giants compete in a “moligopoly” where entry into each other’s (tipped) monopoly origins market is the exception, and competition takes place between them through “indirect entry” into untipped markets.[8] This is competition based on “exploration” of new products, ideas, markets, demand to discover or to create, rather than “rivalry” within a given market as traditionally understood in antitrust. Petit’s theory appears to be a reflection of the observation that on markets with network effects and increasing returns to adoption, competition takes places for the market rather than in the market. When applied to digital markets, it suggests that competition takes place between the tech giants in markets that have not yet been won (ie tipped) by one tech giant or the other and mainly involves competition for indirect entry into untipped markets where there is potential demand. Petit suggests that the tech giants do compete in this way with each other for (new) markets that have not yet tipped, and the type of competition approach and enforcement one adopts should depend on whether the allegedly anticompetitive conduct concerns tipped or untipped markets. Petit suggests a heavier enforcement hand in tipped markets and a lighter touch in untipped markets.[9]

Aspects of Petit’s proposal strike right at the heart of antitrust laws, by raising the question of what competition means and what the purpose of competition law is. As basic and as fundamental these questions are, they have not been resolved, and their answers will determine one’s approach to more or less every question on whether a given business conduct violates the law or not.[10] If rivalry is not what competition means, then what does competition mean and what is the point of competition law? Yet, even these questions beg the further question of what “rivalry” means in the first place. The main problem that makes existing competition laws unfit for digital markets appears to be an understanding of rivalry as the multiplicity of firms competing in the same market for the same group of consumers, which – as Petit demonstrates – may not be how competition operates on these markets.

Petit’s prescription is one with which enforcers and policymakers should seriously engage. This is so because it forces enforcers and policymakers to consider where their intervention can have the highest societal impact at the lowest cost. Government efforts focused on “rivalry spirited remedies” such as break ups, interoperability or data-sharing measures that aim to move tipped markets to a different market structure by generating rivalry are costly and could be futile.[11] If a market has already tipped due to network effects and increasing returns to adoption on the demand side, injecting rivalry into such market structures should not be the aim of enforcers, according to Petit: rather they should aim to promote “a pressure equivalent to” rivalry.[12] For Petit, such an approach involves “maintaining pressure on monopoly rents” by which “antitrust creates an uncertainty equivalent to rivalry that produces powerful incentives on established big tech firms to invent new products and introduce market-shifting innovations”.[13]

There is nothing in Petit’s proposals concerning the limitations of “rivalry” as an antitrust principle that would go against competition rules as they exist on paper: no major competition law prohibits monopoly or dominance per se. This latter fact shows, in itself, that when it comes to the exercise of unilateral market power the law’s preoccupation is not with the number of firms in a given market or the particular structure of that market. If it was, any creation or existence of a monopoly position would have been prohibited in itself.[14] In contrast, the law is concerned with how that market power is exercised. It is only the abuse or anticompetitive exercise of substantial market power that is prohibited under competition laws. Yet, the concept of abuse and the features of anticompetitive conduct are open to interpretation and have been interpreted differently across jurisdictions and over time. In theory, it makes absolute sense to apply competition law in such a way in digital markets that the actual nature of competition between players on such markets is acknowledged and factored in. And, this may well mean that “rivalry” as we know it should not be the guiding principle of enforcement, as Petit suggests. The real challenge to Petit’s proposals, thus, comes not from theory, but from practice.

The difficulty with Petit’s framework lies in the question of how to implement his proposals that competition law should be used to maintain pressure on monopoly rents and that antitrust remedies should aim to promote uncertainty in big tech firms’ monopoly markets that have tipped.[15] According to Petit, “[t]he main function of antitrust in digital markets should not be to promote rivalry— it is often socially inefficient— but a pressure equivalent to it”.[16] Some of the ways in which he suggests that this can be done include direct or indirect control of the firms’ profits, price levels or structures, or share of output in markets that have tipped.[17] Although these appear to follow intuitively from Petit’s general framework, as acknowledged in the book, they pose several challenges. Some of these challenges are potentially formidable.

The challenges include the traditional limitations of direct antitrust intervention into price levels or price structures (which essentially involves regulation rather than antitrust). These limitations also encompass the practical and theoretical challenges of determining what the optimal level of output by a monopolist in a tipped market is. As Petit suggests, indirect ways of applying pressure on the market share of the monopolist in a tipped market might take the form of allowing horizontal mergers by third parties (eg between those on the paying side of a multisided platform). This might limit the exercise of market power by the incumbent. There could also be more promising avenues to explore.[18] Yet, this also threads a fine line because it might mean that a monopoly could be created in one market to combat a monopoly in another market. This could lead to a war of the titans, so to say. This is unlikely to be a pill that competition enforcers can swallow.

Petit’s book grapples with the multidimensional challenges of some very complex market realities. The book is commendable for its acknowledgement of the limitations of some of the proposals that result from that complexity. In any case, none of that takes away from the importance of Petit’s efforts at establishing a conceptual framework that would significantly improve upon the existing competition law framework in digital markets. Thus, even if there may be serious difficulties with the implementation of some of Petit’s proposals, competition enforcers and policymakers should take heed of them. This is because the demonstration of how competition actually works and fails to work in digital markets and the conceptual framework developed in the book have great potential to move us beyond simplistic descriptions of inevitable monopoly tendencies in digital markets. They can guide enforcement of competition law in the right direction by carefully illuminating where the real problems lie and what the right questions to ask, actually, are.

---

* Professor of Law, Director at the Jean Monnet Centre of Excellence in Digital Governance, University of Leeds. The author gratefully acknowledges support by the Leverhulme Trust through the award of a Philip Leverhulme Prize.

[1] Nicolas Petit Big Tech and the Digital Economy (forthcoming, OUP, 2020). The title of this contribution is inspired by “Maria” sung by the nuns in The Sound of Music (1965). For the lyrics which explain the inspiration, see https://genius.com/Richard-rodgers-maria-lyrics.

[2] Petit (n 1) 174 et seq.

[3] Petit (n 1) 174.

[4] Petit (n 1) 169.

[5] Petit (n 1) 169.

[6] Petit (n 1) 169.

[7] Petit (n 1) 169. The second implication, according to Petit, is that “antitrust should adopt tools that allow fact finders to draw a better line between tipped and untipped markets, complementing inferences of monopoly power drawn from structural methods of market definition and evaluation of market power”; ibid.

[8] Petit (n 1) 164. The distinction between “direct entry” and “indirect entry” that Petit adopts is that “direct entry” refers to the supply of perfect substitutes in existing product markets, whereas “indirect entry” refers to the supply of imperfect substitutes or perfect complements in an existing product markets; ibid 157.

[9] Petit (n 1) 184-185, 186.

[10] For the current author’s take on the question of the objectives of competition law, see eg P Akman ‘Searching for the Long-Lost Soul of Article 82EC’ (2009) 29 (2) Oxford Journal of Legal Studies 267 and P Akman The Concept of Abuse in EU Competition Law: Law and Economic Approaches (Hart Publishing, 2012; reprinted, 2015).

[11] Petit (n 1) 175.

[12] Petit (n 1) 184.

[13] Petit (n 1) 184.

[14] For a further discussion of this point in the context of the legislative intent behind the EU competition rules, see Akman (2009) (n 10).

[15] Petit (n 1) 184.

[16] Petit (n 1) 184.

[17] Petit (n 1) 187 et seq.

[18] See Petit (n 1) 193.